We often think that in the many relationships in society where we depend on the services of others it is always better to have more perfect agents. But consider three cases:
A. Imagine you own a company, but you can’t manage it all by yourself. You hire someone to run it for you – an agent — who says they are great at their job. But here’s the catch: that person might not always prioritize your interests. They might, for example, pursue their own goals instead of yours, lack information on what you want them to do, lack the skills needed, be lazy, or even be corrupt and divert the company’s profits into their pockets. And because you hired them to do a job that you can’t do, you can’t immediately detect that they are not doing that job properly.
When you discover that company outcomes are below expectations, you become frustrated and announce new rules that require employees to come to work every day instead of working remotely, so you can observe that they are indeed at work — even if you can’t tell if they are working optimally. That is, you police what you can see, not what may be causing the shortfall. When you still find company performance outcomes lower than hoped for, you decide that the costs of correcting them exceed the net gains from perfect performance. So, you leave it alone, and accept imperfect performance by your agents.
B. Imagine that you are expanding your company into a new technology area. You go into the job market and advertise for employees with special skills in this exciting new, changing area. People with such skills that exactly fit what your company needs now are scarce, and expensive. You then realize that it is counterproductive for your company to hire such perfectly prepared employees – such perfect agents.
So you save on the costs of searching and of paying top market prices for talent and instead hire people who have relatively high skill levels but are highly adaptable, and you have your existing managers train them as needed, changing their skills mix to meet changing competitive needs. You end up better off than you would have been, had you paid for perfect agents. That is, you prefer to hire imperfect agents.
C. Your small technology company is hacked over a weekend, shutting down servers that are badly needed by your customers. You have highly-skilled, long-time employees who are frequently recruited by competitors. Their stock options have already vested; they can leave at any time. They know that you and the company as a whole are desperately dependent on getting the servers back online. Instead of putting up with the extreme pressure of fixing it, yet again, they can simply jump ship.
Instead, they work all weekend – and feel they really had no choice but to do that. People in agency roles who perceive that the principal they serve is highly dependent on their performance – for example, may experience preventable harm — can exhibit the fiduciary norm. This social norm instructs selfless behavior on the part of agents. It makes agents more perfect at no cost to the principal.
In these scenarios, understanding the “Theory of Agency” sheds light on how relationships in business and in society between principals and agents, or people who make decisions on behalf of others, are managed in practical ways. We might expect that we would always prefer to seek to have perfect agents acting in our interests, but in real world situations we often – indeed, we normally – do not receive such performance, and often even prefer to not pay the costs of achieving it. But we may also experience the benefit of social norms expected in people acting as agents under conditions of dependency.
June 2024 marks the 50th anniversary year of this theory proposed by Barry Mitnick, professor of Organizations & Entrepreneurship and of Public and International Affairs at the University of Pittsburgh’s Joseph M. Katz Graduate School of Business and College of Business Administration.
Mitnick and financial economist Stephen Ross, then also at Penn, independently coined the approach’s name, the “theory of agency.” Both had 1973 papers introducing the theory of agency. Mitnick’s 1974 doctoral dissertation at the University of Pennsylvania, written over 1972-1974 (and the 1973 paper drawn from it), first developed the institutional theory of agency. It identified key logics in the behaviors of agents and applied them to diverse organizational and institutional settings. Ross’s work introduced the economic theory of agency, which offered formal economic models of incentive systems in agency.
In 1976, University of Rochester economists Michael Jensen and William Meckling developed an economic agency model of the firm which proved enormously influential over the years, generating tens of thousands of citations. The theory of agency has become so familiar and commonly used in social science that articles sometimes omit citations to the originating works.
Impactful Researcher and Dedicated Teacher
Mitnick, who joined the Pitt Business faculty in 1978, has won numerous awards, including the 2021 Chancellor’s Distinguished Teaching Award. His business history course, Market Manipulations, was selected by the Aspen Institute in 2019 as one of the top ten courses in the world in business and society.
Mitnick’s career spans decades of impactful research, dedicated teaching, and numerous publications. As a respected scholar and mentor, Mitnick’s contributions continue to inspire new generations of business leaders and academics. For example, on SSRN, a prominent website where social scientists from around the world post their papers, the number of times his research has been downloaded puts him in the upper 0.21 percent (two-tenths of one percent) of all social scientists. In addition, Mitnick’s essays on The Conversation, a site where academics explain their research to the public, have been read 125,950 times, according to that website’s tracking data.
Mitnick tells the story of the Jeep Grand Wagoneer he and his wife owned many years ago, when their children were small and they wanted a big, heavy, safe car to transport the family. The Jeep went back to the dealer 50 times during the year they owned it, usually for multiple repairs. “I had a big file of all the repair orders,” Mitnick says. The dealer never returned the car with all the work correctly done. But, as long as the outstanding repairs did not involve a safety issue, and the family needed the car on the road and not in the shop, Mitnick picked up the car – and brought it back again when something else went wrong. The costs of obtaining a perfect repair (with an unavailable car) exceeded the costs of driving a rattling vehicle. After all, agency is rarely perfect….